MECHANICAL SEPARATION 311 grease fresh water is added and the temperature brought to 160°F. It then passes to the secondary centrifuge where the last trace of water is removed. The scouring liquors will average about 1 Ib. of grease to 80 Ib. of liquor treated, there is consequently recovered about 1,500 Ib. of grease per day of 10 hr. or for a 300-day year 450,000 Ib. which at 14 cts. a pound is worth $63,000. Some of the laboratory separations which may be effected and some suggestive of possible commercial separations are: bacteria from water, emulsions and serums, pigment from paint, fractional separation of clay particles, clarification of fruit juices, clarification of varnishes and glues, and the separation of emulsified oils from an unemulsified portion. SEPARATION OF SOLIDS FROM GASES Dust Chambers.—In theory, with dust laden air at ordinary temperatures a dust chamber of large cross-section and relatively short length should be as effective as a long chamber of small cross-section. With hot fumes the long chamber is preferable because of lowering of temperature with consequent reduction in volume and velocity and with increase of moisture in the non-gaseous form. Reduction of temperature on the other hand tends to increase the adsorption of the solid particles and increases the density of the air but these factors in decreasing the rate of settlement are not of much moment. If the fumes are hot but non-corrosive the wall of the settlement chambers should be as thin as possible. In one zinc-oxide dust-collection system the dust-laden air is cooled by passing it through pipes of thin copper. The exigencies of fume collection in smelting are such that only massive chambers of brick work are found to be satisfactory. In lead-smelting operations it is necessary to have fairly high velocities in the dust chambers in order to maintain the pressure sufficient to force the fume through the filtering bags of the baghouse. In this case settlement of dust in the chambers is not so important as collection of fume and it may be generally stated that when the particles are too small to settle in any dust chamber and require a filtration or positive means of separation after dust-chamber passage the extent of the dust chamber becomes of secondary importance. In the case of lead smelting it is the great volume of the dust and air and its high temperature which requires extensive settlement chambers. Much material is obtained by settlement but the most valuable material in point of assay comes from the baghouse. If the dust which is collected comes from cold currents and only the very fine particles are of commercial value then a settlement chamber of limited cubical contents is in order. If the volume of gases is small and it is merely desired to get rid of a dust nuisance, dust chambers need not be installed and the separatory arrangements will resolve themselves into some form of separatory or knock down devices. Tortuous chambers and chambers arranged with baffles should be avoided as they tend to back up the draft.1 It must be evident that in order to move the dust and fume either by artificial draft or by natural draft there must be a certain velocity and pressure at the point where the dust originates. If obstruction be employed in the dust chamber the pressure and velocity at the point where the dust originates may be 1 Much better than baffles are plates hung with axes parallel to the flow of the gas. Wires suspended in the chamber are also more valuable than are the transverse baffles, but after all it is questionable if anything is an improvement upon, making the dust chamber large enough to slow down the gas current to less than 5 ft. per sec., and for true fume nothing can replace the bag house or Cottrell installation-—EDITOR.